• Everyday AI
  • Posts
  • Ep 776: OpenAI’s Codex For Beginners: Why it’s better than Claude, ChatGPT and How to Start Using It

Ep 776: OpenAI’s Codex For Beginners: Why it’s better than Claude, ChatGPT and How to Start Using It

Google just launched its Gemini Deep Research Agent, Microsoft unveiled a 100-agent AI cybersecurity system called MDASH, and Meta rolls out private AI chat, and more

Author

Everyday AI
May 13, 2026

 In Partnership With Adobe  

Adobe just introduced an entirely new way to create with Firefly AI Assistant. 

Adobe just introduced an entirely new way to create — bringing the power and precision of its creative suite into one conversational experience. With Firefly AI Assistant — now live in the Adobe Firefly app, the all-in-one creative AI studio - just describe what you want to create, and the assistant orchestrates the workflows, drawing on 60+ pro-grade tools across Adobe Creative Cloud apps, including Photoshop, Illustrator, Premiere, Lightroom and more — to help bring your ideas to life.

Every step the assistant takes is visible, so you can refine, redirect, or take over at any time. You stay in control of the outcome as the creative director. Check it out today at firefly.adobe.com.

Outsmart The Future

Today in Everyday AI
8 minute read

🎙 Daily Podcast Episode: OpenAI’s Codex is quickly turning into much more than a coding tool, with AI agents that can control apps, browse the web, and automate real desktop workflows from one place. Give today’s show a watch/read/listen.

🕵️‍♂️ Fresh Finds: Claude just doubled coding and API limits, Google is testing an AI-powered screen pointer, and Android Auto is getting smarter with new AI features and screen support, and more. Read on for Fresh Finds.

🗞 Byte Sized Daily AI News: Google just launched its Gemini Deep Research Agent, Microsoft unveiled a 100-agent AI cybersecurity system called MDASH, and Meta rolls out private AI chat, and more. Read on for Byte Sized News.

💪 Leverage AI: Codex is the undisputed best AI system out there. We show you how beginners can get started. Keep reading for that!

↩️ Don’t miss out: Miss our last newsletter? We covered: Google is launching its Gemini Omni video agent, Google releases Gemini Intelligence and AI hardware, and OpenAI unveiled a new AI cyber initiative called Daybreak, and more. Check it here!

Ep 776: OpenAI’s Codex For Beginners: Why it’s better than Claude, ChatGPT and How to Start Using It


OpenAI's Codex is the best AI system on the planet. 🪐

Yet, most non-technical users have never touched it.

Don't let the word 'code' in Codex scare you. Codex is OpenAI's forthcoming superapp, and if you haven't used it yet, you're literally asking for your competition to lap you.

On today's 'AI at Work on Wednesdays' we're going over the basics: What Codex is, how to use it, and why it's actually better than ChatGPT and Claude.

Tune in LIVE to find out.

Also on the pod today:

• Codex vs Claude memory showdown 🧠 
• Uptime: Codex crushes Claude 🚦 
• Live demo: Codex desktop powers 💻

Listen on our site:

Click to listen

Subscribe and listen on your favorite podcast platform

Listen on:

Here’s our favorite AI finds from across the web:

New AI Tool Spotlight –  Memoket Gem captures every important conversation and turns it into context your AI tools can actually use, APIDeck Connects Claude, Codex, Cursor, and any MCP client to Accounting, CRM, HRIS, and 4 more Unified APIs, Blaze sets your strategy, creates your content, and gets smarter every week

Claude Code Updates — Claude just doubled code and API limits, plus dropped smart new agent features to speed up AI development.

Krea 2 Release — Krea 2 lets you control style in AI image creation and even mix up to four style references at once.

Claude for Legal Work — Claude just rolled out 20+ new connectors and 12 plugins made for legal work, letting lawyers plug AI directly into the software they already use.

Paypal AI — PayPal showed that letting employees pick their own AI tools sped up trust and adoption across the company.

Mouse Pointer AI — Google is testing an AI-powered pointer that lets you interact with anything on your screen, just by pointing and speaking.

Android Auto Upgrades — Android Auto is getting a big upgrade to fit wild new car screen shapes and add features like YouTube streaming and smarter AI help.

Anthropic Acquires Stainless — Anthropic is in talks to buy Stainless, a key dev tools startup used by OpenAI and Google, for $300 million.

Rivian Assistant Release — Rivian Assistant puts you in control of your car with smart, natural voice commands.

AI Powered Radio — thehype just launched a nonstop AI-powered radio station for founders and builders, streaming real-time analysis and industry signals.

1. Google Debuts Deep Research Agent for AI-Powered Analysis 🤓

Google has just rolled out its Gemini Deep Research Agent, a new tool that autonomously tackles complex research tasks using web search, code execution, and document analysis.

This upgrade means AI can now deliver detailed, cited reports on topics ranging from market trends to technical comparisons, saving users hours of manual digging. The agent works asynchronously, allowing for collaborative planning and even generating charts or visualizations when asked—perfect for those who need more than a quick summary.

2. Google I/O Preview: Android’s Quick Share Gets AirDrop Integration ☔

In a timely move ahead of Google I/O, Google announced that Android’s Quick Share will soon work seamlessly with Apple’s AirDrop, bridging a long-standing gap between the two platforms.

Users will be able to wirelessly transfer photos, apps, messages, and eSIM data across Android and iOS devices from top brands. The update also brings QR code sharing for even faster cross-platform transfers from any Android phone.

3. Microsoft Unleashes AI Agents to Hunt Windows Vulnerabilities 🛡️

Microsoft revealed its new AI-powered cyber defense system, codenamed MDASH, which just helped researchers uncover 16 new vulnerabilities in the Windows networking and authentication stack, including several critical remote code execution flaws.

The big news is that this new system orchestrates more than 100 specialized AI agents to discover, debate, and even prove security bugs at enterprise scale, outperforming traditional single-model approaches and topping public vulnerability benchmarks.

4. Meta unveils encrypted Incognito Chat for AI with no message logs 🤫

Meta says its newly announced Incognito Chat for Meta AI is designed so conversations are not stored at all, including on Meta’s own servers, marking a notable shift in how consumer AI tools handle privacy. CEO Mark Zuckerberg says the feature uses end-to-end encryption, setting it apart from rival “temporary” chat modes that still retain logs for days or longer.

The timing is significant as AI companies face mounting lawsuits and public scrutiny over stored chat data tied to violent incidents and media disputes. Meta says the feature will roll out gradually to WhatsApp and the Meta AI app over the coming months.

5. Claude follows Codex’s lead and gets a goal feature 🥅

Anthropic has rolled out a new /goal command in Claude Code that lets developers set a clear completion condition and have Claude keep working across turns until it’s met, marking a notable step toward more hands-off AI workflows.

The update matters now because it shifts Claude from a reactive helper into something closer to a persistent worker that can finish tasks without constant prompting. In simple terms, users tell Claude what “done” looks like, and a separate evaluator checks progress after every turn until the job is finished or canceled.

Anthropic pretty much invented the agentic desktop, then got lapped on their own field.

How?

While everyone obsessed over Claude Code and Claude Cowork, OpenAI quietly shipped Codex on the desktop in February and it's been eating their lunch since. Most leaders kept scrolling because the name starts with "code," and you know that voice. "code" in the name so not for me.

Biiiig mistake, fam.

What's the big deal with Codex? It's the first true autonomous agent that actually works WITH you in real time, using your apps and files while you keep doing your own thing.

We're calling it the biggest shift in knowledge work since ChatGPT first dropped.

Don't worry. We've got your Codex 101 for the nontechnical leader.

On today's Everyday AI, we break down three unlocks that put Codex measurably ahead of Claude Code, Claude Cowork, and ChatGPT, plus why your company is prolly leaking money.

You've gotta window. Let's get after it.

1. The Memory Unlock 🔥

Most leaders assume Claude on the desktop remembers what they told it yesterday, and we hate to break the news, but nope.

Anthropic toggles a memory setting in the desktop app that does absolutely nothing. Tell it your favorite NFL team in one chat, ask in the next, and it shrugs at you like a stranger.

Same problem hits Claude Cowork and Claude Code, three little silos that have zero idea the others exist.

Codex runs ONE unified workspace where memory actually persists across every project and chat, forever.

Here's the gnarly part.

Your team has prolly been pouring context into Claude for months thinking it sticks, and that knowledge has been quietly evaporating.

Try This

Open Codex tonight and feed it three things about how your business actually runs. Your top product, your biggest customer segment, your weekly reporting rhythm.

Start a fresh chat tomorrow and ask Codex to draft your Monday update using everything it knows. The difference will smack you in three seconds, and that's the moment you stop renting amnesia and start building real leverage.

2. The Uptime Unlock ⚡

Enterprise software lives or dies on three nines, which is 99.9 percent uptime and the gold standard every serious tool gets measured against.

Codex sits pretty at a clean 100, while Claude Code hangs back at 99.2. Sounds like a rounding error, but wayyy bigger than it looks.

Stack three dependent automations and one missed trigger snaps the whole chain like a cheap necklace. Now run that across 20 workflows every weekday.

Your morning briefing fails to send, your CRM enrichment skips a Tuesday, and your sales digest ships half built before anyone catches it Wednesday.

This is the silent tax leaders never see until quarterly numbers come in soft.

Try This

Audit every automation you've spun up in any AI tool over the past 90 days, and flag the ones touching revenue, customers, or executive reporting.

Move those over to Codex first this week, then block one quiet hour Friday to migrate the rest while it's still fresh. Your future self pulling a clean Monday dashboard will very quietly thank you.

3. The Execution Unlock 🚀

Most agents hijack your computer the moment they start working, leaving you on your hands like a kid in detention.

It runs its own cursor at the operating system level with its own mouse, keyboard, and browser. So while you're heads down on a deck, it's quietly triaging your inbox, and while you're on a call it's pulling reports across five apps without bothering you.

Two workers, one machine. Sooo much output.

This is the first time the agent stops riding shotgun and becomes a true parallel coworker, which means the real unlock isn't speed, it's YOUR highest-leverage hours getting handed back on a silver platter.

Leaders who figure this out in the next 30 days will run at 2X capacity by February.

Try This

Pick one repeatable task that quietly eats 30 minutes a day. Inbox triage, calendar prep, that weekly report you grind out by hand.

Hand it to Codex with three rounds of feedback, then tomorrow morning let it run while you focus on work that moves the needle. Once you feel parallel execution kick in, there's truly no going back.

Time to move.

 🚨 BONUS 🚨

Make sure to repost today’s show on LinkedIn to skip the line for our upcoming Cowork Codex Cohort.

Reply

or to participate.